The “brain in a vat” has long been a staple of philosophical thought experiments and science fiction. Now scientists are one step closer to creating the real thing, which could enable groundbreaking experiments of a much more empirical kind. Research teams at Stanford University and the RIKEN Center for Developmental Biology in Japan have each discovered methods for coaxing human stem cells to form three-dimensional neural structures that display activity associated with that of an adult brain.

By applying a variety of chemical growth factors, the RIKEN researchers transformed human embryonic stem cells into neurons that self-organized in patterns unique to the cerebellum, a region of the brain that coordinates movement. The Stanford team worked with induced pluripotent stem cells derived from skin cells and chemically nudged them to become neurons that spontaneously wired up into networks of 3-D circuits, much like the ones found in the cerebral cortex—the wrinkled gray matter of the brain that supports attention, memory and self-awareness in humans.

“For years people have used mouse embryonic stem cells to generate teratomas—things that look like they could be organs,” says David Panchision, a neuroscientist at the National Institutes of Health, which supported the Stanford research. “But it's not organized and systematic, the way a developing brain needs to be to function.” In contrast, the Stanford team's neural structures not only self-assembled as cortexlike tissue, the neurons also sent signals to one another in coordinated patterns—just as they would in a brain. The cerebellar tissue generated by the Japanese scientists did, too.

So what could one do with a working chunk of lab-grown brain? Using it to someday grow neural spare parts for diseased or aging patients “is not impossible,” says RIKEN's Keiko Muguruma. But the near-term goal is to subject these living mini brains, dubbed “organoids” by scientists, to medical research that is otherwise impossible or unethical. “You can do detailed, mechanistic experiments that are directly relevant to human disease,” Panchision explains. “If you're looking for very specific molecular targets or pathways in the brain, and how drugs might act on them, the difference between human cells and mouse cells is significant.”

Panchision foresees organoids being used in virtual clinical safety trials for new psychiatric medications. “Most brain disorders aren't understood at the circuit level,” he says. So whereas growing spare parts for your brain remains a fantasy for now, having these neural crash-test dummies for research purposes could be the next best thing.